When I used Coot to refine a dimer protein structure, I got a problem. The chain A of dimer protein possesses whole density for cofactor NADP+. But the NADP+ in chain B misses the density in nicotinamide, the other portion possess density. So, in this case, how to refine the NADP+ in chain B? Thank you for your suggestions.
Dear Xinxin,
I see many more problems with your refinement. My simple advice is to remove the NADP in question and refine everything else till you get almost no difference densities anywhere else besides the site you investigate, i.e. eliminate all red and green densities anywhere else besides the site in question. A clear positive (green) electron density should emerge. Why I said that there are many other problems? Because a lot of other side chains have no ideal densities as well there are some subtle conformational errors. Create as ideal as possible stereochemistry, refine solvent structure (water molecules and other solvents). Only then the phases fidelity can allow you to decide what you have at the other side. I am not even mentioning a plethora of other problems with the data that can impede the model refinement, like a proper space group, appropriate Rmerg, sufficient completeness, desired isotropy of reflections. Without examining the statistics of data collection it is literally impossible to advise you on how to proceed. In my hands most of the PDB structures have smaller or larger errors. DO not be fulled into the trap of barely passable structure. For more advise you would have to find my old posts on desired characteristics of well done crystal structure.
Good Luck.
Bog
- Thank Professor Stec very much. And I will to learn more knowledge from your old posts. Thank you again for your advice
- Badges
- Science method